Yes in wicket 1.3 there is only 1 active page per pagemap But it shouldn't matter where the page comes from. If it is serialized from disk or serialized in the session somehow by the container, You still should be able to fully construct all your data again in the objects.
So what is suddenly null? All the transient fields? Why aren't those reloaded from a database or something? johan On Nov 27, 2007 5:01 PM, Evan Chooly <[EMAIL PROTECTED]> wrote: > We have a page that presents a report in which we can "drill down" into > the > data. When this happens, we setResponsePage() back to a new instance of > the > page with some different parameters and that all works dandy. But when we > hit the back button, it seems the old version is gone. All the model data > comes back null (though the model object itself is not). So when we click > back and then click another link to drilldown in another direction we get > NPEs. From stepping through DiskPageStore, et. al, it would appear that > since the page class isn't changing, the new page bumps the old one from > the > cache so that only one version of the page. > > Am I correctly interpreting what I'm seeing and is there a way to set the > number of versions of a page? >
